[Use of C response protein in predicting postoperative anastomotic leakage in patients with rectal cancer].

OBJECTIVE: To investigate the value and feasibility of C reactive protein (CRP) in predicting postoperative anastomotic leakage in rectal cancer patients with enhanced recovery after surgery (ERAS) for safer implementation of this ERAS. METHODS: A cohort study on serum CRP of 455 rectal cancer patients undergoing laparoscopic radical resection according to the ERAS procedure at Gastrointestinal Unit of General Surgery Department, Guangdong General Hospital from August 2014 to June 2017 was retrospectively carried out. The serum CRP level was measured before operation and at postoperative days 1-7, and the serum CRP level of the groups with and without anastomotic leakage was compared to analyze its prediction for anastomotic leakage. Diagnostic standard of anastomotic leakage was based on the definition of postoperative anastomotic leakage in rectal cancer from International Study Group of Rectal Cancer (ISREC): (1) Postoperative localized or diffuse peritonitis occurred, or fecal liquid was found from the abdominal drainage tube; (2) When anastomotic leakage was uncertain, peritoneal or pelvic computed tomography scan should be used to confirm. RESULTS: All the 455 patients underwent surgery successfully, and 41 patients (9.0%) had anastomotic leakage postoperatively. Patients with anastomotic leakage were diagnosed (4.0±2.0) days postoperatively, of whom 8 cases (19.5%) were diagnosed more than 5 days postoperatively. Serum CRP levels in patients with anastomotic leakage continued to increase within 1-4 days postoperatively [(50.04±27.98) mg/L to (122.75±52.98) mg/L] and decreased 5 days postoperatively [(92.02±58.26) mg/L], both were higher than those of non-anastomotic leakage group, and the difference was statistically significant (all P<0.05, except postoperative day 2). The serum CRP level of non-anastomotic leakage group reached the peak on the second postoperative day [(83.10±37.45) mg/L] and decreased 3 days postoperatively [(48.01±27.59) mg/L]. The ROC curve was drawn with the anastomotic leakage as the state variable, and the CRP level as the detection variable. The area under the curve (AUC) at postoperative 1, 2, 4, 5, 6 and 7 days was 0.74, 0.58, 0.83, 0.82, 0.65, and 0.70, respectively. The maximum was at postoperative day 3 [0.93(95%CI: 0.86-0.99)]. The Youden index was 0.72, and the threshold of CRP was 80.09 mg/L, as the cut-off point to predict anastomotic leakage, with sensitivity, specificity, and positive predictive value of 79.3%, 92.3%, and 74.2%, respectively. CONCLUSIONS: Monitoring the postoperative serum CRP level can help predict the occurrence of anastomotic leakage after laparoscopic surgery for rectal cancer. When the serum CRP level is >80.09 mg/L on the third postoperative day, the CRP level has the largest value in predicting postoperative anastomotic leakage, and the safety of ERAS has a certain clinical significance as well.